Meta and Google Ads: what cannot be compared in a report

The table that reassures and lies

The request comes up at every review: a table with Google Ads on the left, Meta on the right, and a total column. It's visually satisfying, and statistically wrong from the conversions row onward.

The problem isn't a settings issue. The two platforms don't answer the same question. Google Ads reports how many conversions followed an interaction with its ads, by its rules. Meta reports the same thing by its own. Neither knows the other exists.

Three columns that don't mean the same thing

Conversions. Meta attributes on seven days after click and one day after view by default. Someone sees an Instagram ad, doesn't click, then buys the next day after a Google search: Meta counts the sale, Google counts it too. The sum of both exceeds the real number of sales.

**Conversion rate**. The denominator changes. Google Ads calculates on clicks, Meta generally against landing page views. Two rates both showing 3% don't describe the same reality.

Conversion value. It depends on what each platform receives from the site. If Meta's ROAS includes tax and Google's doesn't, the gap has nothing to do with performance.

What compares safely

Four families of figures stay honest side by side.

Spend, because it comes from billing rather than a model. **Impressions** and clicks, with one caveat: on Meta, take link clicks rather than all clicks. **CPC**, bearing in mind that a Google click and a Meta click carry different intent.

And above all, everything you measure yourself downstream: orders in the back office, leads in the CRM, revenue in the accounts. Those figures belong to no platform, so they can't be counted twice.

A presentation that holds up

Rather than a three-column table with a false total, structure the report in two parts.

First the overall result, measured outside the platforms: number of sales or leads, revenue, average acquisition cost derived from total spend divided by that real result. That block is accurate and defensible.

Then the platform breakdown, presented as such: what each one claims, with its attribution window stated plainly. No total column.

This structure has an unexpected benefit: it makes the conversation simpler. The client understands that a platform figure is a claim, not an accounting truth, and stops asking why the totals don't reconcile.

Automating without repeating the mistake

Most reporting tools aggregate sources without warning what they're adding together. That's the trap when connecting several platforms: Adwise reports each source with its attribution window, and flags the cases where a total would mislead rather than computing it anyway.